Beruflich Dokumente
Kultur Dokumente
SC34-6459-03
CICS Transaction Server for z/OS
SC34-6459-03
Note!
Before using this information and the product it supports, be sure to read the general information under “Notices” on page
139.
Summary of changes . . . . . . . . . . . . . . . . . . . . . . ix
Changes made to this book for CICS Transaction Server for z/OS, Version 3
Release 1 . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Changes made to this book for CICS Transaction Server for z/OS, Version 2
Release 3 . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Changes made to this book for CICS Transaction Server for z/OS, Version 2
Release 2 . . . . . . . . . . . . . . . . . . . . . . . . . . ix
New and changed function in CICS Transaction Server for OS/390 Release 3 x
Dynamic routing . . . . . . . . . . . . . . . . . . . . . . . . x
CICS business transaction services . . . . . . . . . . . . . . . . xi
Temporary storage queue models . . . . . . . . . . . . . . . . . xi
Long temporary storage queue names . . . . . . . . . . . . . . . xii
Resource Recovery Services (RRS) . . . . . . . . . . . . . . . . xii
Sysplex enqueue models . . . . . . . . . . . . . . . . . . . . xii
Support of IIOP inbound to Java applications . . . . . . . . . . . . . xiii
Coupling facility data tables . . . . . . . . . . . . . . . . . . . xiii
Support for CICS Web enhancements . . . . . . . . . . . . . . . xiii
Java Virtual Machine . . . . . . . . . . . . . . . . . . . . . xiv
Open Transaction Environment . . . . . . . . . . . . . . . . . . xiv
FEPI . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv
Support for Tivoli Global Enterprise Manager . . . . . . . . . . . . . xiv
Changes made to this book for CICS Transaction Server for OS/390 Release 3 xv
Contents v
Prerequisite hardware . . . . . . . . . . . . . . . . . . . . . 92
Storage requirements . . . . . . . . . . . . . . . . . . . . . 92
CICSPlex SM naming conventions . . . . . . . . . . . . . . . . . 92
Generic names . . . . . . . . . . . . . . . . . . . . . . . . 93
Security planning . . . . . . . . . . . . . . . . . . . . . . . . 93
Protecting access to CICSPlex SM functions and CICS resources . . . . . 94
Special considerations for BAS . . . . . . . . . . . . . . . . . . 94
CICS command and resource checking . . . . . . . . . . . . . . . 94
Defining time zones . . . . . . . . . . . . . . . . . . . . . . . 94
An overview of the installation process . . . . . . . . . . . . . . . . 95
Preparing to set up a CAS . . . . . . . . . . . . . . . . . . . . 95
Preparing to set up a CMAS . . . . . . . . . . . . . . . . . . . . 96
Using the batched repository-update facility . . . . . . . . . . . . . 97
Preparing to set up a MAS . . . . . . . . . . . . . . . . . . . . 98
Reuse of CICSPlex SM object definitions . . . . . . . . . . . . . . . 98
Where next? . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . 127
The CICS Transaction Server for z/OS library . . . . . . . . . . . . . 127
The entitlement set . . . . . . . . . . . . . . . . . . . . . . 127
PDF-only books . . . . . . . . . . . . . . . . . . . . . . . 127
Other CICS books . . . . . . . . . . . . . . . . . . . . . . . 129
Determining if a publication is current . . . . . . . . . . . . . . . . 129
Accessibility . . . . . . . . . . . . . . . . . . . . . . . . . 131
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Trademarks. . . . . . . . . . . . . . . . . . . . . . . . . . 140
Notes on terminology
Unless otherwise indicated, in the text of this book:
v The term CICSPlex SM (spelled with an uppercase letter ‘P’) means the IBM®
CICSPlex System Manager element of CICS Transaction Server. The term
CICSplex (spelled with a lowercase letter ‘p’) means the largest set of CICS
systems to be managed by CICSPlex SM as a single entity.
v The term CICS system refers to a single administration unit of CICS on any of
the platforms managed by CICSPlex SM. On MVS™, this term is synonymous
with CICS region.
v The term MVS refers to the MVS/Enterprise Systems Architecture (MVS/ESA)
operating system.
Changes made since the last edition are marked by vertical bars in the left margin.
Changes made to this book for CICS Transaction Server for z/OS,
Version 3 Release 1
CICSPlex SM support for the CICS for Windows component of IBM TXSeries™
(also known as NT 4.3 and NT 5.0) is no longer provided in CICS Transaction
Server for z/OS, Version 3 Release 1. Therefore, it is no longer possible to set up a
CICSPlex SM remote MAS agent for Windows.
However, customers, who wish to do so, can continue to use the CICS Transaction
Servers Version 2.3 or Version 2.2 for CICSPlex SM support of TXSeries.
Changes made to this book for CICS Transaction Server for z/OS,
Version 2 Release 3
CICSPlex SM support for Link3270 bridge workload management is introduced in
this release. It is covered in Chapter 6, “Managing workloads,” on page 57.
CICS for OS/2 is not supported as an RMAS in CICS Transaction Server for z/OS,
Version 2 Release 3.
A CICSPlex SM CMAS will now run only on a CICS system at the same release
level. The ability to run the latest release of CICSPlex SM on older releases of
CICS in the CMAS has been removed. Similarly, a MAS being used as a CICS
Transaction Server for z/OS, Version 2 Release 3 CICSPlex SM Web User
Interface server must also be at the same CICS release level as the Web User
Interface.
Changes made to this book for CICS Transaction Server for z/OS,
Version 2 Release 2
The following change has been made to this book:
v Information on CICSPlex SM support for directly or indirectly connectable
CICS/VSE systems from CICS Transaction Server for z/OS, Version 3 Release 1
is removed.
v Information on CICSPlex SM support for Enterprise JavaBean technology.
There has been a change in CICSPlex SM field naming conventions in this release.
Data set name fields such as DSNAME, file name fields such as LOCFILE and
REMFILE, and transient data queue name fields such as EXTRATDQ and
INTRATDQ are now case-sensitive. When entering data set and file names into the
CICSPlex SM interfaces (end user interface, API and the web user interface),
ensure that you enter the data in the correct case. In previous releases of CICSPlex
SM, the data set names and file names are automatically converted to upper case.
Dynamic routing
CICS Transaction Server for OS/390 Release 3 introduces extensions to the
CICSPlex SM dynamic routing program (called the dynamic transaction routing
program in previous releases). In addition to dynamic routing of transactions
initiated at a terminal, you can now dynamically route:
v Transactions invoked using the EXEC CICS START commands that are
associated with a terminal
v Transactions invoked using the EXEC CICS START commands that are not
associated with a terminal
v Distributed program links, including:
– The CICS Web support
– The CICS Transaction Gateway
– EXCI calls
– CICS Client ECI calls
– Distributed Computing Environment (DCE) remote procedure calls (RPCs)
– Open Network Computing (ONC) remote procedure calls (RPCs)
– Internet Inter-Object Request Block Protocol (IIOP)
– Any function that issues an EXEC CICS LINK PROGRAM request
v Transactions associated with CICS business transaction services (BTS) activities
(see “CICS business transaction services” on page xi)
These new functions allow you to integrate workload management for requests from
all these sources and thereby improve performance and workload throughput. You
specify which transactions and programs may be dynamically routed, and the CICS
regions in which they may be run.
BTS processes are CICSplex-enabled, and take full advantage of CICSPlex SM’s
workload separation and workload balancing functions. The CICS system group
within a CICSplex, across which related BTS processes and activities may run, is
known as a BTS-set. When an activity is attached, WLM routes the activity to the
most appropriate region in the CICSplex. You can separate a BTS workload based
on process-type. Workload management and the CICS Transaction Affinities Utility
understand affinities between BTS processes and activities.
Summary of changes xi
The existing temporary storage operations views, TSQ, TSQS, TSQGBL, and
TSQGBLS, remain unchanged. However, you can now delete temporary storage
queues from the TSQ, TSQSHR and TSQNAME views by entering the command
DEL. A new TSQ Deletion Panel asks you to confirm the deletion.
To provide greater flexibility, the length of temporary storage queue names has
been increased to a maximum of 16 characters. For example, users can generate
queue names of the form:
XXXXaUUUUUUUU
where:
XXXX is the transaction id.
a is a suffix allowing the transaction or user to have more than one queue.
UUUUUUUU
is the user id.
The TST or temporary storage model prefix can then be XXXX or XXXXa.
FEPI
CICS Transaction Server for OS/390 Release 3 introduces changes to the way in
which the CICS/ESA Front End Programming Interface (FEPI) is managed using
CICSPlex SM. The changes are:
v New BAS views:
– FENODDEF, which defines FEPI nodes
– FEPOODEF, which defines FEPI pools
– FEPRODEF, which defines FEPI property sets
– FETRGDEF, which defines FEPI targets.
v FEPI resources are now installed using these new BAS views; they cannot be
installed using the operations views
Summary of changes xv
xvi CICS TS for z/OS: CICSPlex SM Concepts and Planning
Chapter 1. Introducing CICSPlex SM
The CICSPlex System Manager element of CICS Transaction Server for z/OS,
Version 3 Release 1 (CICSPlex SM) is a system management tool that enables
you to manage multiple CICS systems from a single control point. Enterprises in
which CICSPlex SM may be needed range from those running 10 or 15 CICS
systems, to those running two or three hundred (or more) CICS systems: in the
latest MVS sysplex environment, having such large numbers of CICS systems to
support a transaction-processing workload is becoming increasingly common.
For this release of CICSPlex SM, the connectable CICS systems are:
| v CICS Transaction Server for z/OS 3.1
v CICS Transaction Server for z/OS 2.3
v CICS Transaction Server for z/OS 2.2
v CICS Transaction Server for OS/390 1.3
| You can use this release of CICSPlex SM to control systems running supported
| releases of CICS that are connected to, and managed by, your previous release of
CICSPlex SM. However, if you have any directly-connectable release levels of
CICS, as listed above, that are connected to a previous release of CICSPlex SM,
you are strongly recommended to migrate them to the current release of
CICSPlex SM, to take full advantage of the enhanced management services. See
the CICS Transaction Server for z/OS Migration from CICS TS Version 2.3 for
information on how to do this.
Table 1 shows which supported CICS systems can be directly connected to which
releases of CICSPlex SM.
| Table 1. Directly-connectable CICS systems by CICSPlex SM release
| CICS system CICSPlex SM CICSPlex SM CICSPlex SM CICSPlex SM
| component of CICS component of CICS component of CICS component of CICS
| TS 3.1 TS 2.3 TS 2.2 TS 1.3
| CICS TS 3.1 Yes No No No
| CICS TS 2.3 Yes Yes No No
| CICS TS 2.2 Yes Yes Yes No
| CICS TS 1.3 Yes Yes Yes Yes
| TXSeries 4.3.0.4 No Yes Yes No
| TXSeries 5.0 No Yes Yes No
|
Features of CICSPlex SM
This section describes briefly the system management functions of CICSPlex SM:
v A real-time, single-system image (SSI)
CICSPlex SM provides a real-time, single-system image (SSI) of all CICS
regions and resources that make up your enterprise’s transaction processing
environment. CICSPlex SM creates an inventory of the CICS systems and
resources, and maps the relationships between them; this is called the
CICSPlex SM topology. Thus the operator no longer needs to know the location
of a resource before working with it. The SSI is provided by the CICSPlex SM
operating function and applies to all CICSPlex SM applications: operations,
monitoring, real-time analysis (RTA), and workload management (WLM).
v A single point of control
The CICSPlex SM operator can manage all of the enterprise CICS systems and
their resources from a single session. That is, CICSPlex SM can provide a single
point of control for the enterprise, which in turn means that the CICSPlex SM
operator is able to manage large and complex configurations of CICS resources.
Furthermore, you can have multiple points of control, so that multiple CICS
Where next?
Chapter 2, “The CICSPlex SM environment,” on page 7 describes the configuration
and topology of CICSPlex SM and the system management functions in more
detail. Chapter 3, “Designing your CICSPlex SM environment,” on page 17 starts
guiding you through the process of designing your own CICSPlex SM environment.
A Starter Set is supplied that contains all you need to get started. See Chapter 10,
“The CICSPlex SM Starter Set,” on page 101 for the details, including the
CICSPlex SM and CICS resource definitions. You might find it useful to refer to
these for examples as you read the following chapters.
Overview
This section gives you a brief overview of the main elements of the CICSPlex SM
environment. See Figure 1 on page 8.
| The MASs in a CICSplex can be managed by several CMASs, but only one is
| defined as the maintenance point CMAS responsible for keeping the data used by
| each CMAS synchronized.
CMASs communicate across CMAS-to-CMAS links, which are typically used for
routing management commands and data between CMASs. A coordinating address
space (CAS) is used to set up the CICSPlex SM component topology, and to
support the MVS/TSO ISPF end-user interface (EUI) to CICSPlex SM.
Web TSO
Browser Terminal
MVS/TSO
ISPF
CAS-to-CAS Link
Data
Repository CAS
CMAS-to-CMAS Link
CMAS
Web User
Interface
Server
MAS
Agent Code
MAS
CICSplex
CICSplex
WUI
Repository
Figure 1. Key components of a CICSPlex SM configuration. Each of these components is described in this chapter.
The CICSplex
For CICSPlex SM’s purposes, a CICSplex is any grouping of CICS systems that
you want to manage and manipulate as if they were a single entity. That is, a
CICSplex is a management domain, made up of those CICS systems for which you
want to establish a single system image (SSI). A CICSplex managed by
CICSPlex SM could include every CICS system in your enterprise. Alternatively,
you could define multiple CICSplexes, each of which would include a logical
grouping of CICS systems. For example, a CICSplex could comprise all CICS
systems on a particular MVS image, or all CICS systems accessible by a subset of
your users, or all CICS systems serving a particular geographical area.
Furthermore, the composition of a CICSplex can be altered without affecting the
functions of the underlying CICS systems. The CICS systems in a single CICSplex
managed by CICSPlex SM do not have to be explicitly connected to each other for
management purposes.
You can assign an unlimited number of CICS systems and CICS system groups to
an existing CICSplex.
Although you can define a CICS system to only one CICSplex, you can assign a
CICS system to multiple CICS system groups within the CICSplex. You can also
assign the CICS system group to any number of other CICS system groups.
All the MASs within a CICSplex are managed by the same CICSPlex SM Address
Space (CMAS) or the same group of CMASs.
data
repository
CMAS
Agent code
MAS
(CICS system)
MVS
Figure 2. MASs and their CMAS. Each of these components is described in this section.
The MASs within a single CICSplex do not have to be explicitly connected to each
other for CICSPlex SM-specific communication. However, CICS connections
required in support of transaction routing and function shipping are still required.
MASs can be defined and managed as individual CICS systems or grouped into
system groups within the CICSplex. Each system group is one or more CICS
systems that you want to manage as a unit subset of the CICSplex. System groups
are described in “Identifying system groups” on page 21.
In CICS Transaction Server for z/OS, Version 3 Release 1 all MASs are local, that
is they run on the same MVS image as the CMAS by which they are managed. You
do not need to define an explicit link between the CMAS and the local MAS. For
local MASs, system-management data is accumulated in data-space caches and is
communicated to the CMAS via the Environment Services System Services (ESSS)
address space (which is described on page “Environment Services System Services
(ESSS)” on page 11).
| Remote MASs running on a different MVS image from the managing CMAS or on
| other operating systems such as Windows or OS/2 are not supported in CICS
| Transaction Server for z/OS, Version 3 Release 1 or later.
A MAS can be set up as a CICSPlex SM Web User Interface server. In this case,
the CICS release level of the MAS and the CICSPlex SM Web User Interface
server must be the same.
The CMAS implements the BAS, WLM, RTA, and monitoring functions of
CICSPlex SM, and maintains configuration information about the CICSplexes it is
managing. It also holds information about its own links with other CMASs. It stores
this information in its data repository. (See Figure 2 on page 10).
Note: The CMAS does not support user applications or terminals, and its resources
should not be considered available for non-CMAS tasks, including (without
limitation) the use of any monitoring and performance tool other than those
supplied as a part of CICSPlex SM.
A CMAS cannot be running a lower release of CICS than its MASs. Also, both the
CMAS and the MASs must be running the same release of CICSPlex SM.
A CMAS is not part of any CICSplex: a single CMAS can participate in the
management of multiple CICSplexes, but belongs to none of them.
CICSPlex SM can issue SNA generic alerts to NetView, provided that a CMAS is
installed on the same MVS image as the NetView instance with which
CICSPlex SM is working.
# There is one instance of ESSS for each version or level of CMAS on any MVS
# image.
The server component of the Web User Interface runs on a dedicated CICS
Transaction Server for z/OS, Version 3 Release 1 CICSPlex SM local MAS
connected to a CMAS at the same release level.
The Web User Interface is supplied with a set of linked menus and views but can
also be customized to reflect your business procedures and to suit the needs of
individual users.
| The general principles of the WUI are described in CICSPlex System Manager Web
| User Interface Guide.
# A CAS need not be started on every MVS system where a CMAS runs to provide
# Single System Image (SSI) function for EUI requests directed at a CICSplex.
# However in order for an EUI user to set the context to a CMAS, that CMAS must be
# connected to an active CAS.
The general principles of the CICSPlex SM user interface are defined in the
CICSPlex System Manager User Interface Guide.
Note: When you are viewing or updating CMAS configuration data, the context
is a CMAS rather than a CICSplex, and any scope value is ignored.
(Remember that the CICSplex is the largest single unit on which any
CICSPlex SM command can act.)
v The scope is always a subset of the context: its purpose is to limit the effects of
CICSPlex SM commands to particular CICS systems or resources. Indeed, the
effects of any command cannot be felt beyond the current scope. The scope can
be a single CICS system, a CICS system group (made up of any number of
CICS systems in the CICSplex), or a business application (represented by a
resource description).
You can look at CICS systems or resources anywhere in your enterprise from one
point, the single point of control, simply by changing the context and scope.
CICSPlex SM objects
To define the configuration of your CICS systems to CICSPlex SM, and to define
your BAS, WLM, RTA, and monitoring requirements, you create CICSPlex SM
objects, and associate them with each other. For each object, and for each
association or link between them, a record is created in a CICSPlex SM data
repository. Figure 3 on page 14 shows how the CICSPlex SM objects relate to
each other.
Resource
description
Resource Resource
group definition
Resource
CICSplex assignment
MVS
RTA, WLM, RTA, WLM, RTA, WLM,
or MON or MON or MON
specification group definition
CICSPlex SM definitions and CICS resource definitions held on the data repository
can be managed in the following ways:
v Using the EUI views; see “The CICSPlex SM end user interface (EUI)” on page
12
v Using the batched repository-update facility; see “The batched repository-update
facility”
v Using the API; see the CICSPlex System Manager Application Programming
Reference and CICSPlex System Manager Application Programming Guide.
You can create a visual map of the definitions in your data repository. The map can
be of business application services, resource monitoring, real-time analysis, or
workload monitoring. You select a starting point for the map, which might be, for
example, a CICS system group or an individual resource, and CICSPlex SM
displays the starting point and all the definitions that either refer to that definition or
are referred to by it. You use either the MAP view command or the batched
repository-update facility to create a map.
The batched repository-update facility is also useful for migrating your definitions
from one platform to another. You can use it to retrieve existing resource definitions
from the CICSPlex SM data repository then, after making any required changes to
the definitions, input the changes to another batched repository-update facility run
that creates resource definitions on the new platform.
Tivoli Business Systems Manager uses the CICSPlex SM API, EXEC CPSM, to
discover the CICSPlex SM objects it monitors. Tivoli Business Systems Manager
collects events generated as a result of SAM (Systems Automation Monitoring) and
RTA (Real Time Analysis) processing, in order to monitor the status of these
resources.
Tivoli Business Systems Manager also provides the user with the ability to link to
the CICSPlex SM Web User Interface.
For details of the Tivoli Business Systems Manager agent for CICSPlex SM, see
the Tivoli Business Systems Manager CICSPlex SM Release Notes.
Where next?
Chapter 1, “Introducing CICSPlex SM,” on page 1 and this chapter have given you
an overview of CICSPlex SM and the facilities it offers to help you administer your
enterprise. The rest of this book tells you how to design your CICSPlex SM
environment and how to start planning for and defining those functions you wish to
use. The chapters point you to where you can find the details of how to perform the
required tasks.
This chapter helps you to design your CICSPlex SM environment by giving you
guidance on each step of the process. An example is used to illustrate the process.
Figure 4 on page 18 shows an example of the type of map you should be aiming to
produce.
TOR 1 TOR 2
CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3
AOR 4 CICS/400 3.1
AOR 1 AOR 3 CICS Transaction Server
CICS Transaction Server CICS Transaction Server for OS/390 2.3
for OS/390 1.3 for OS/390 1.3 CICS/400 3.1
AOR 5
AOR 2 CICS Transaction Server
FOR 1
CICS Transaction Server for OS/390 2.3
CICS Transaction Server
for OS/390 1.3 for OS/390 1.3
TOR 3
AOR 8
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 6
AOR 9
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 7
CICS Transaction Server
for OS/390 1.3
System E System F
Figure 4. A map of the enterprise CICS systems. This enterprise has 15 CICS systems distributed among OS/390,
z/OS, and OS/400, environments.
Your initial map of the enterprise CICS systems should include every operating
environment in which CICS is installed. It should also show:
v The current version and release of the operating systems being used
v The CICS systems running in each environment, and the CICS platform, version,
and release of each one
v The resource-manager role of each CICS system, if applicable
In the example map shown in Figure 4 on page 18, only the CICS systems running
under OS/400® cannot be managed by CICSPlex SM. On your own map, choose a
convention, such as shading or use of color, for marking CICS systems that cannot
be managed by CICSPlex SM. However, don’t remove them from the map
altogether. If you decide to move those systems to a CICS platform or release that
CICSPlex SM can manage, they can be reinstated easily in the enterprise map.
If you do not plan to use workload management facilities, there are no restrictions
on how you combine CICS systems and CICS system groups to form a CICSplex.
For example, you might associate CICS systems by:
v Geographic area within the CICSplex.
v Function, such as all CICS systems that are application-owning regions (AORs),
file-owning regions (FORs), or terminal-owning regions (TORs).
v Application, such as the CICS systems serving as AORs, FORs, and TORs that
are used by a specific application or group of applications.
v Time period, such as all CICS systems that are normally active during specific
hours of the day or night.
If you do plan to use workload management facilities, you must ensure that:
v Each CICS system that is to act as a target for specific workload processing
must be in the same CICSplex as the CICS systems acting as routing regions.
(The routing regions and target regions associated with a CICSplex can be in the
same or different MVS images.)
v CICS systems acting as the routing regions must be CICS TS systems.
The question is, how do you decide what to do? There are no hard-and-fast rules
governing the number of CICSplexes you define, but there are some guidelines that
will help you select the most suitable configuration for your enterprise. These
guidelines are presented in the form of a three-step process:
1. Decide to have a single CICSplex for the entire enterprise
2. Look for reasons why a single CICSplex might not be feasible
3. Confirm your decision
Group 8 in CICSplex 1 and Group 7 in CICSplex 2 include the same set of CICS
systems as the CICSplex to which it belongs. These are often useful groups to
define because the scope value (as specified for a monitor specification, for
example) can be a CICS system or a CICS system group name only: it cannot be
the name of a CICSplex.
This is merely an initial list of system groups. It is likely to be added to (or altered)
when BAS, WLM, RTA, and monitoring requirements are identified.
You can identify one or more subsets of the CICS systems within a CICSplex as a
CICS system group, which can be manipulated as a single entity and independently
of the rest of CICSplex. For example, if you define a CICSplex made up of TOR,
AOR, and FOR CICS systems, you might want to define the AORs as a CICS
system group, so that you can use a single CICSPlex SM command to make
changes to, or request data from, all CICS systems in that category.
CICSplex 1
TOR 1 TOR 2
CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3
AOR 1 AOR 3 AOR 4 CICS/400 3.1
CICS Transaction Server CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3 for OS/390 2.3
CICS/400 3.1
AOR 2 FOR 1 AOR 5
CICS Transaction Server CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3 for OS/390 2.3
CICSplex 2
TOR 3
AOR 8
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 6
AOR 9
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 7
CICS Transaction Server
for OS/390 1.3
System E System F
Figure 5. Identifying the CICSplexes. The example enterprise operates as two discrete units: the first three MVS
images process a different workload from the remaining two, and there is no sharing of resources between those two
groups. Accordingly, the enterprise CICS systems are to be divided between two CICSplexes. CICS systems that
cannot be managed by CICSPlex SM (the CICS/400 systems, in the example) have been shaded.
Alternatively, you could define a single group for any of the following:
v Particularly heavily loaded CICS systems
v CICS systems that have different security requirements from the other CICS
systems in the CICSplex
v CICS systems in which particular applications run
Any duplication of CICS system names that occurs in this way (for example, if a
particular CICS system belongs to more than one constituent group) is
accommodated by CICSPlex SM. When a CICS system group is the target of a
CICSPlex SM command, CICS systems appearing in the group more than once are
acted on once only.
Locating CMASs
| The CMAS is the component of the CICSPlex SM topology that is responsible for
| most of the work involved in managing and reporting on systems and their
| resources. It is the CMAS that is responsible for presenting the SSI to the operator.
| Each CICSplex is managed by at least one CMAS. This section gives you guidance
| on deciding where to put CMASs and how many to have.
You should also consider the performance implications of your decision. Each
CMAS has its space requirements and its own data repository. If you plan to use
the ISPF EUI on the image, you will need to install a CAS. For details of estimated
space requirements of the CICSPlex SM components, see Chapter 9, “Planning
CICSPlex SM installation and setup,” on page 91.
Figure 6 on page 26 shows the example map updated again, this time to show the
application of these CMAS rules and recommendations.
CMAS-to-CMAS links
When a CICSplex crosses multiple MVS images, multiple CMASs are likely to be
involved in managing that CICSplex. The CMASs must be connected to each other
(using CICS intercommunication methods) so that data relating to the CICS
systems belonging to the CICSplex is accessible by all relevant CMASs, and an
SSI can be presented to the CICS operator. The minimum requirement is that
CMASs managing the same CICSplexes be linked to form an open chain of
CMASs, so that each CMAS is linked, if only indirectly, to every other CMAS.
Meeting this requirement ensures that there is at least one path from each CMAS to
every other CMAS involved in managing the same CICSplex. To achieve the best
performance and availability, link every CMAS directly to every other CMAS.
The links between multiple CMASs involved in managing the same CICSplex are
responsible for the SSI of the CICS systems or systems that belong to that
CICSplex. However, you do have a degree of choice regarding the number of links
you define. First, here’s a reminder of the rules and recommendations concerning
CMAS-to-CMAS links:
v To establish an SSI of a CICSplex, the CMASs managing that CICSplex should
at the very least be linked to form an open chain.
CICSplex 1
TOR 1 TOR 2
CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3
AOR 1 AOR 3 AOR 4 CICS/400 3.1
CICS Transaction Server CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3 for OS/390 2.3
CICS/400 3.1
AOR 2 FOR 1 AOR 5
CICS Transaction Server CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3 for OS/390 2.3
MP
CMAS 4
CMAS 5
CICSplex 2
TOR 3
AOR 8
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 6
AOR 9
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 7
CICS Transaction Server
for OS/390 1.3
System E System F
Figure 6. Adding CMASs to the enterprise map. A CMAS has been installed on each MVS image. CMAS 3 can issue
SNA generic alerts to the NetView instance on the same MVS image. CMAS 3 is the maintenance point CMAS for
CICSplex 1, and CMAS 5 is the maintenance point CMAS for CICSplex 2. (That is, CMAS 3 will be the context CMAS
when you define CICSplex 1, and CMAS 5 will be the context CMAS when you define CICSplex 2 to CICSPlex SM.)
TOR 1 TOR 2
AOR 1 AOR 3
AOR 2 AOR 4
CICSplex
Figure 7. CMAS-to-CMAS links. This CICSplex is made up of six CICS systems, three managed by CMAS1 and three
managed by CMAS5. In the absence of a direct link between the two CMASs, CICSPlex SM is able to navigate
dynamically around its “network” to gather information relevant to the CICSplex. It could go via CMAS4 or, if CMAS4 is
unavailable, for example, it could go via CMAS2 and CMAS3. For optimum performance, however, a direct link should
be added between CMAS1 and CMAS5.
v The more links you establish, the better the performance. The best performance
is achieved when every CMAS is directly connected to every other CMAS
involved in the management of a single CICSplex.
v A multiplicity of CMAS-to-CMAS links is also better for availability: if CMAS 1 is
at the end of the chain and is connected to CMAS 2 only, the failure of CMAS 2
breaks the chain and leaves no connection to CMAS 1.
CICSplex 1 CICSplex 2
CMAS1 to CMAS2 CMAS4 to CMAS5
CMAS1 to CMAS3 CMAS5 to CMAS4
CMAS2 to CMAS1
CMAS2 to CMAS3
CMAS3 to CMAS2
CMAS3 to CMAS1
You could have decided not to link CMAS 1 to CMAS 3 directly. This arrangement
would have met the minimum CMAS-to-CMAS link requirement (by leaving us with
an open chain of CMASs) and would have reduced the number of links to be
defined by two. Had you done so, information requested from System C about
CICS resources on System A, for example, would be obtained indirectly (via the
When there are multiple CMASs on the same MVS system, managing the same
CICSplex, a local MAS that specifies only the CICSPLEX(name) CICSPlex SM
system parameter, and not the CMASSYSID(name) CICSPlex SM system
parameter, will connect to the last CMAS to initialize on the MVS image that
manages the named CICSplex for a specific release of CICSPlex SM. See the
CICS Transaction Server for z/OS Installation Guide for a description of
CICSPlex SM system parameters.
Alternatively, data belonging to one CICSplex can be accessed from an MVS image
on which a second CICSplex resides via CAS-to-CAS links. With the configuration
you have chosen, CICSPlex SM information about an AOR on System E, for
example, requested from the CICSPlex SM user interface on System A, can be
obtained via CAS-to-CAS links rather than CMAS-to-CMAS links. CAS-to-CAS links
are discussed in “CAS-to-CAS links” on page 31.
The maintenance point CMAS is the CMAS that is the context value when the
CICSplex is defined to CICSPlex SM. (The context is described on page “Display
formats” on page 12.) Each CICSplex can have only one maintenance point CMAS,
but a single CMAS can be the maintenance point for more than one CICSplex. If
the maintenance point CMAS is unavailable at any time, changes to all
BAS does not support temporary maintenance point CMASs. The maintenance
point CMAS must be available if you are creating or maintaining BAS definitions.
CICSplex 1
TOR 1 TOR 2
CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3
AOR 1 AOR 3 AOR 4 CICS/400 3.1
CICS Transaction Server CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3 for OS/390 2.3
CICS/400 3.1
AOR 2 FOR 1 AOR 5
CICS Transaction Server CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3 for OS/390 2.3
MP
CMAS 4
CMAS 5
CICSplex 2
TOR 3
AOR 8
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 6
AOR 9
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 7
CICS Transaction Server
for OS/390 1.3
System E System F
Figure 8. Adding CMAS-to-CMAS links to the example map. Eight links have been defined, giving full connectivity
among CMASs managing the same CICSplexes.
Note: CICSPlex SM’s single point of control means that changes to the
maintenance point CMAS can be made from any system on which the
CICSPlex SM EUI is available.
Locating CASs
Once you have decided where to install CMASs and how to link them, you can turn
your attention to the CASs.
CMAS 1 CMAS 2
Figure 9. Valid CAS configurations. The most likely configuration is that shown on System A.
However, those shown on System B and System C are also valid. CICSPlex SM data cannot
be accessed from System B because there is no MVS/TSO instance. However, because
CAS2 has links to other CASs on MVS images where MVS/TSO is available, the data on
System B is accessible from other MVS images. System C has no CMAS, but can still be
used as an access point for CICSPlex SM data on other MVS images.
| It’s very easy to update the example map by applying these rules: there are five
| CMASs, so there must be five CASs, one on each of the MVS images on which a
| CMAS is installed. Figure 10 on page 32 shows the addition of the CASs..
CAS-to-CAS links
If you define no CAS-to-CAS links, the CICSPlex SM operator can sign on to
CICSPlex SM via CAS 1 (or CAS 2, or CAS 3) and still have a complete view of
the CICS systems and resources in CICSplex 1 because the CMASs managing
CICSplex 1 are linked to each other. The same observation is true of CICSplex 2.
For example, a CICSPlex SM operator accessing CICSPlex SM via CAS 5 can
see data for TOR 3 via the links between CMAS 4 and CMAS 5. However, for both
performance and availability reasons, it is better to establish links between CASs
also.
CICSplex 1
TOR 1 TOR 2
CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3
AOR 1 AOR 3 AOR 4 CICS/400 3.1
CICS Transaction Server CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3 for OS/390 2.3
CICS/400 3.1
AOR 2 FOR 1 AOR 5
CICS Transaction Server CICS Transaction Server CICS Transaction Server
for OS/390 1.3 for OS/390 1.3 for OS/390 2.3
CAS 4 CAS 5
MP
CMAS 4
CMAS 5
CICSplex 2
TOR 3
AOR 8
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 6
AOR 9
CICS Transaction Server
CICS Transaction Server
for z/OS 2.2
for z/OS 2.3
AOR 7
CICS Transaction Server
for OS/390 1.3
System E System F
|
| Figure 10. Adding CASs to the example map. A CAS has been added to every MVS image to which a CMAS has
| been assigned.
In the example, few CAS-to-CAS links are essential. If you add two links, one from
CAS 1 to CAS 4, and one from CAS 4 to CAS 1, both CAS 1 and CAS 4 can
function as an single point of control because each has access to data about
CICSplexes 1 and 2. However, for maximum performance and availability, and
because the total number of CASs in the enterprise is small, you are going to link
every CAS to every other CAS. As with the CMAS-to-CMAS links, the formula for
calculating the number of links to be defined is n² − n, where n is the number of
CASs. In the example, there are five CASs, and so 20 links must be defined to
establish full connectivity for the enterprise. If you have room on your map, add the
CAS-to-CAS links. As the example map is beginning to look crowded, the
CAS-to-CAS links are listed separately in Table 2.
Table 2. CAS-to-CAS links required for full connectivity among the example CASs
CAS 1 to: CAS 2 to: CAS 3 to: CAS 4 to: CAS 5 to:
CAS2 CAS1 CAS1 CAS1 CAS1
CAS3 CAS3 CAS2 CAS2 CAS2
CAS4 CAS4 CAS4 CAS3 CAS3
CAS5 CAS5 CAS5 CAS5 CAS4
Full connectivity among the enterprise CASs is recommended for performance and
availability reasons. However, it is not essential: if the enterprise CMASs are
adequately connected to each other, a single point of control for the enterprise can
be established without full connectivity among the CASs.
If you opt for the gradual approach, you will need to revisit your final map of the
enterprise CICS systems and identify those you plan to start with. For example, if a
staged implementation were adopted for your example map, you could decide to
begin by implementing CICSplex 2 only. The CAS-to-CAS links and the
CMAS-to-CMAS links would be much reduced in number (two CAS-to-CAS links
rather than 20, and two CMAS-to-CMAS links, rather than 12), but the other
decisions made about CICSplex 2 are still valid. That is, the same CICS system
groups can be defined, and the CAS and CMAS installations are still required.
Where next?
Now that you have a map of your enterprise, you can start the planning for the
installation of CICSPlex SM and the creation of your proposed configuration. See
Chapter 9, “Planning CICSPlex SM installation and setup,” on page 91 for guidance
on planning for the installation of CICSPlex SM.
Defining CAS-to-CAS links is one step in the process of installing a CAS, which
forms part of the setup procedure. This is described in detail in CICSPlex System
Manager Administration.
To enter this information, you can use the appropriate CMAS Configuration
Administration (ADMCONFG) views, or the batched repository-update facility.
CICSPlex SM also provides the following views that you can use to manage CMAS
configuration definitions when the associated CMASs are active.
CICSPLEX
To display information about the CMASs associated with a CICSplex known
to the local CMAS.
CMAS
To display information about CMASs known to the local CMAS, shut down a
CMAS, and alter CMAS component trace settings.
CMASD
To display information about a specific CMAS known to the local CMAS,
shut down a CMAS, and alter CMAS component trace settings.
CMASPLEX
To display information about the CICSplexes being managed by the local
CMAS.
CMTCMLNK
To display information about all or specific CMASs linked to the local CMAS
and to discard a CMAS-to-CMAS link.
CMTCMLND
To display information about the link between a specific CMAS and the local
CMAS and to discard the CMAS-to-CMAS link.
CMTPMLNK
To display information about all or specific MASs linked to the local CMAS
and to discard the CMAS-to-MAS link.
CMTPMLND
To display information about a specific MAS linked to the local CMAS and
to discard the CMAS-to-MAS link.
See CICSPlex System Manager Administration for details of all these views.
To enter this information, you can use the appropriate CMAS Configuration
Administration (ADMCONFG) views, or the batched repository-update facility.
When you have created a CICSplex definition, you can associate CICS systems
(MASs) and CICS system groups with the CICSplex. See “Preparing CICS system
definitions.”
To create CICS system definitions, and to display information about CICS systems
defined to a CICSplex, you use the CICSSYS view.
For every CICS system that is to be managed by CICSPlex SM, you must gather
the following information:
v The name by which this system is to be known to CICSPlex SM
v The name of the CICSplex to which the CICS system is to belong
v The APPLID, that is, the VTAM application identifier for the CICS system
v The SYSID, that is, the CICS system identifier
This is sufficient data for adding the MAS definition to the CICSPlex SM data
repository. If you know that you have special BAS, WLM, RTA, or monitoring
requirements for any CICS system, you can enter those at the same time.
Alternatively, you can update the CICS system definition later as new requirements
become apparent. For example, you might want to collect some of the following
information, which is optional and can be supplied after the CICS system is defined
to CICSPlex SM:
v If system availability monitoring (SAM) is to be implemented, the name of the
primary CMAS, which is the CMAS to which the CICS system usually connects,
and the hours of operation of the CICS system
v Whether CICS command checking is to be simulated
v Whether CICS resource checking is to be simulated
v The time zone of the CICS system, if it is not to be the same as that of its CMAS
To enter this information, you can use the Topology Administration (ADMTOPOL)
views, or the batched repository-update facility. Note that the ADMTOPOL views are
also used to create time-period definitions.
Defining CICS system groups, perhaps more than any other part of the design
stage, is an iterative process and a cumulative one. It might be obvious to you
already that there are natural groupings of CICS systems within your CICSplexes,
and that you will want to manipulate each of those groups as a single entity. If this
is the case, start to document the groups now. Because groups do not have to be
mutually exclusive, and because a CICS system can belong to any number of
groups, you can add CICS system group definitions as new requirements emerge.
You can display information about the CICS system groups, and the CICS systems
associated with them, that are known to the CICSplex identified as the current
context, by using the SYSGRPC view.
To create CICS system link definitions, and to display information about those
definitions, you use the SYSLINK view.
CICSSYS
System ID
Model System
Create System Links
CONNDEF
Connection Attributes
SYSLINK
SYSLINK
To create and administer the links between CICS systems in the CICSplex.
This view is described in CICSPlex System Manager Managing Business
Applications.
CICSPlex SM provides the MAS view that you can use to manage configuration
definitions when the associated CICS systems are active.
MAS To manage information about active CICS systems known to a CICSplex or
connected to the designated CMAS involved in managing the CICSplex.
v Display information about the active CICS systems
v Browse or temporarily update the general, WLM, real-time analysis,
resource monitoring, and MAS trace attributes of an active CICS system.
v Stop MAS agent code in an active CICS system.
This view is described in CICSPlex System Manager Administration.
Note: Using the CICSSYS and PERIODEF views to update definitions affects both
a currently running system and the definitions in the data repository. Updates
made using the MAS view, however, are not included in the data repository.
Where next?
You have a working CICSPlex SM environment, with your configuration and
topology defined, and with the links between the entities defined. The entities can
communicate with each other and you can run transactions. However, to optimize
the operation of your enterprise and to take full advantage of the facilities offered by
CICSPlex SM, you need to set up your BAS, WLM, RTA, and monitoring
requirements. For guidance, go to Chapter 5, “Managing resources using Business
Application Services (BAS),” on page 41, Chapter 6, “Managing workloads,” on
page 57, Chapter 7, “Exception reporting using real-time analysis (RTA),” on page
73, or Chapter 8, “Collecting statistics using CICSPlex SM monitoring,” on page 83,
as appropriate.
Limitations of RDO
RDO is the traditional CICS method of defining resources to CICS systems. With
RDO, resource definitions are tied to a single group and groups that are processed
sequentially from a group list. Any duplicate definition found later in the process
overrides any earlier one. In addition, the very nature of the definitions that RDO
requires and the extent to which they are available to multiple CICS systems can
prove limiting. RDO has several disadvantages in a CICSPlex SM environment:
v Resources are assigned to groups that are assigned to specified CICS systems.
v Resource definitions have to be duplicated across CICS systems, unless the
CSD is shared.
v Each end of every communications link has to be defined explicitly.
BAS concepts
BAS enables you to view and manage your resources in terms of their use in your
business, and independently of their physical location. This section introduces the
basic concepts of the BAS function.
Logical scoping
Once your CICS resources are defined to CICSPlex SM, you can monitor and
control resources in terms of their participation in a named business application,
rather than their physical location in the CICSplex. Logically-related resources can
be identified and referred to as a set, regardless of where they actually reside at
any given time.
If you set the scope to be your application, any operation or monitoring views will
display only those resources that satisfy your selection criteria. This gives you the
power to control precisely how those resources are managed.
The CICSPlex SM data repository (EYUDREP) serves as the central repository for
all your CICS resource definitions. This minimizes the number of resource
definitions you need for your CICSplex by:
v Providing a single system image approach to defining CICS resources across all
supported platforms.
v Producing both local and remote instances of a resource from a single definition.
v Managing multiple versions of a definition. For example, you can have different
versions of the resources for an application as it progresses through a number of
test phases.
v Generating multiple CICS communication links from a single set of connection
and session definitions.
| With BAS, communications links are created dynamically from model connection
| and session resource definitions. You define connections and sessions that describe
| the nature of the link. The BAS system link (SYSLINK) object is used to create the
| link between two CICS systems, by specifying the connection and session
| definitions to be used. The connection and session definitions can be used by any
| number of system link definitions that share the same characteristics. For more
| information see CICSPlex System Manager Managing Business Applications and
| CICSPlex System Manager Administration.
You can use either or both of these forms, depending on the situation and degree of
precision you require.
The simplest form is the migration form using resource descriptions. You create
resource definitions and resource groups, and associate them with one or more
resource descriptions. It is the resource description that defines the logical scope
for the resources. The resources are assigned to specific CICS systems and you
therefore still need separate definitions for a resource that is local to one system
and remote to another. This approach to BAS is similar to using RDO, in that the
resource description is analogous to the group list. You in effect create a version of
your CSD on the CICSPlex SM data repository. Though this is a good position from
which to establish your CICSPlex SM resources environment, you cannot take
advantage of all the facilities offered by BAS.
Using resource assignments allows you to manage your individual resources and
change their attributes for individual systems, or to suit special circumstances. You
can:
v Control resources of a given type in a given group.
v Identify resources as either local or remote, and assign them to various CICS
systems with a single resource definition.
v Process selected resources from a group by specifying a filter expression.
v Modify resource attributes for a particular use by specifying override expressions.
RDO BAS
Resource
CSD Group List Description
CSD Group
Resource Group
PROGRAM
PROGRAM TRANSACTION
CICS System
Resource Resource
Assignment Description
Resource Group
PROGRAM TRANSACTION
Managing BAS
Your BAS environment is set up and managed using the objects described in
Table 3.
Table 3. Resource administration objects
Object name Description Function
RASGNDEF Resource assignments A resource assignment describes the characteristics of
the selected resource definition type, and how those
resources are to be assigned to CICS systems. The
assigned resources must all be of one type, for example,
files, and must belong to a resource group. A resource
can be assigned as both local and remote in one or
more CICS systems. Before you can use the resource
assignment, you must associate with a resource
description; see the RESDESC object.
The object models showing the relationships between these objects are shown in
Figure 14 on page 47 and Figure 15 on page 48.
!Autoinstall
!Error handling
CICSGRP
RESDESC
!Logical Scope
!Autoinstall
!Resource Group Scope
RESINDSC
!Groups in Description
RESGROUP
RESINGRP
!Definitions in Group
resDEF
!Resource Attributes
RESGROUP
RESINGRP
Definitions in Group
resDEF
Resource Attributes
The process of moving your resource definitions and resource groups from your
CSDs to the data repository is:
1. Use the EXTRACT command of the DFHCSDUP utility routine to read the CSD
records.
2. Use the CICSPlex SM extract routine EYU9BCSD to create batched
repository-update facility commands.
3. Input these commands to the batched repository-update facility to create
resource definitions and resource groups, and the objects that link them. See
CICSPlex System Manager Administration for more information of the batched
repository-update facility.
Defining resources
BAS enables you to define and maintain your resource definitions by providing a
view for each resource type. See Table 3 on page 45.
Validating resources
BAS performs many of the same resource definitions checks as RDO does. As
individual resources are defined and installed, BAS checks:
v Individual attributes of a resource
v Interdependent resources attributes
v Release-specific resource attributes
In addition, BAS checks that each set of resources is consistent. Every time you
make a change to a resource set, BAS checks that the resource being added or
updated is not in conflict with a resource already in the CICS system. For example,
BAS also provides a MAP function that allows you to check that your resource
associations are what you want. You use this facility to display the structure of your
resource associations, starting at any point in hierarchy. Figure 16 shows an
example of a resource definition map.
There is no real limit to the number or combination of resource definitions that can
make up a group. However, only one version of any given resource can be included
in a resource group at one time. You have to maintain each version of a resource
definition in a different resource group.
When you have migrated or defined your resources and resource groups, you need
to create resource descriptions to define your applications. Resource descriptions
are defined using the RESDESC object. A resource description is directly related to
one or more resource groups that are to be managed as an entity. In the migration
form BAS model, this relationship can be considered to represent a logical scope,
but the resource description is really analogous to a RDO group list, in that all the
resources are tied to a given CICS system or CICS system group.
Note: If you are using the migration form of BAS, you do not use resource
assignments.
Each resource assignment relates to one resource type within a resource group,
and it must be associated with a resource description, if it is to be used in logical
scoping or automatically installed. Each resource group can be in more than one
resource assignment.
You can further refine the selection of resources within the type specified by the
resource assignment by using a filter expression, and you can change the resource
attributes using an override expression. You can combine in the expression an
unlimited number of attribute values, using the logical operators AND, OR, and
NOT. When the resource description is specified as the scope, CICSPlex SM
processes only those resources that meet the specified selection criteria. This gives
you a great deal of control over the management of your resources.
Each resource group to be used as part of the application must be added to the
resource description. You can associate whole resource groups with a resource
description to create a larger set of resources (much like a CSD group list). This
allows you to manage the resources more efficiently. The set of resources identified
in a resource description can be:
v Identified as a logical scope (such as an application) for use in subsequent
CICSPlex SM requests
The BAS objects that are used for creating and managing these resource
associations shown in Figure 14 on page 47 and their functions are summarized in
Table 3 on page 45.
For more information about resource descriptions see the CICSPlex System
Manager Managing Business Applications.
Security considerations
Because of the importance of resource definitions to your CICSPlex SM
environment, you need to consider carefully the implications of allowing users
access to certain types of resource or to certain functions. You can allow a user to
have access to all resource definition views and their related administration views,
or you can restrict access to the definition views for a particular resource type. You
can also define the type of access a user has, for example, one user might have
read-only access to connection definition (CONNDEF) and session definition
(SESSDEF) views, but update access (which permits creation and administration of
resources) to all other resource types. You need to make sure that the BAS views
are adequately protected, so that unauthorized users cannot create and administer
resources.
You should also take care if you are running CICS TS, and are using the EXEC
CICS CREATE command to build new resources. Any definition created with the
This section poses some of the questions you might like to consider when you start
planning to implement BAS in your enterprise:
v If you are going to use the extract facilities:
– In what order do you want to extract your CSDs?
– Do you want all the records from a CSD or a subset?
– Are you going to extract more than one CSD at a time?
v Are you going to use the migration form approach?
v Are you going to implement full-function BAS? You then need to think about how
your business applications use resources and the assignments you need to
create.
v Which resources are to be installed automatically and which are to be installed
dynamically?
v Where are the resources to be installed?
v Do you need to define manually any resources that are not defined in the CSD?
v What security measures do you need to implement? (See “Security
considerations” on page 52).
| The IBM CICS Interdependency Analyzer for z/OS can assist you in the planning
| and understanding of your application resource flow; it shows you:
| v Transaction resource dependencies; that is, the sets of resources used by
| individual CICS transactions, and on which the transactions depend in order to
| run successfully
| v Transaction affinities; that is, those groups of transactions that have affinities with
| each other, meaning that the whole group must be installed in the same region,
| or in a particular region
| For more information about the CICS Interdependency Analyzer, see the CICS
| Interdependency Analyzer for z/OS User's Guide and Reference.
Implementing BAS
This section describes a route you could take to implement BAS on your CICSplex.
There are no resource assignments and, although you can now run your
CICSPlex SM system perfectly satisfactorily, you will not be taking advantage of the
special functionality offered by BAS.
Once you have extracted your CEDA definitions, you can proceed to migrate your
resource definitions to full BAS function, as described in “Migrating to the use of
resource assignments.”
Note: You can use the same resource description to manage both whole resource
groups and selected resources identified in resource assignments. As you
begin to take advantage of the resource assignment capabilities, you may
have to update some of the underlying resource definitions. For example, a
resource definition that was previously associated as is with a CICS system
might require additional attributes before it can be assigned as both a local
and remote resource in different CICS systems.
A recommended approach
When you decide to implement BAS functions in your enterprise, you should begin
by defining the appropriate objects. This section summarizes the preceding
sections. You should:
v Extract resource definitions and resource groups from the CSD, as described in
“Migrating your resource definitions” on page 48. Alternatively:
– Use the appropriate resource definition views to create resource definitions.
– Use the RESGROUP view to create resource groups.
v Use the RASGNDEF view to create resource assignments.
v Use the RESDESC view to create resource descriptions.
Then, to create associations between these objects:
v Use the appropriate resource definition view to add resource definitions to the
resource group.
v Use the RESGROUP view to add the resource group to the resource
assignments.
v Use the RASGNDEF view to associate the resource assignment with a CICS
system or CICS system groups.
You can use the MAP action command to display a visual map of the BAS
definitions in your data repository. If you want to see a list of the objects and
associations you have already defined, use the RESINGRP view.
If you have problems defining any particular object or association, refer to the
CICSPlex SM Starter Set examples. The Starter Set includes BAS examples that
you can use as models for your own objects. You can copy the supplied samples
directly into your data repository via the batched repository-update facility. For more
information about using the Starter Set as a model see the CICS Transaction
Server for z/OS Installation Guide. For more information about the batched
repository-update facility, see CICSPlex System Manager Administration. The
Starter Set definitions are listed in Chapter 10, “The CICSPlex SM Starter Set,” on
page 101.
Where next?
Once you have identified your BAS requirements, you need to set up your
resources and their associations as described in CICSPlex System Manager
Managing Business Applications.
You can specify the location of a resource when you are designing your system.
Then, requests of a specific resource are always routed to the same region.
Typically, the location of the resource is specified in the installed resource definition.
This is known as static routing.
With dynamic routing, the location of the resource is decided at run time. In
CICSPlex SM, the decision on where to run a piece of work is made by the
user-replaceable dynamic routing program (called the dynamic transaction routing
program in previous releases). The user-replaceable program EYU9XLOP creates
the environment necessary for CICSPlex SM-based dynamic routing, and sets up
the run-time environment.
The CICS regions involved in dynamic routing may act as one or more of the
following:
Requesting region
The CICS region in which the work request originates.
Routing region
The CICS region in which the decision is taken on where the work will run.
Target region
The CICS region where the request is actioned.
For dynamic transaction routing, the requesting region and the routing region are
typically TORs, and the target region is typically an AOR.
For inbound DPL client requests, the requesting region and the routing region are
typically TORs, and the target region is typically an AOR.
For EXEC CICS START commands associated with a terminal, the requesting
region is typically an AOR, the routing region is typically a TOR, and the target
region is typically an AOR.
For peer-to-peer DPL requests, EXEC CICS START commands that are not
associated with a terminal, for CICS business transaction services processes and
activities, and for Link3270 bridge requests, the requesting region, routing region,
and target region are typically AORs.
For enterprise bean invocations, the requesting region is typically the external client
code (but can be Enterprise JavaBean code in another CICS region) that invokes
the enterprise bean, the routing region is a CICS listener region, and the target
region is typically an AOR.
The “hub” model is hierarchical, in which routing is controlled by one region (the
routing region, that is, the TOR). Normally, a routing program runs only in the
routing region. This model has the advantage of being relatively simple to
Requesting region
Requesting region
Routing region
Routing region
Target region
Target region
Distributed Distributed
routing routing
program program
Figure 18. Dynamic routing using a distributed routing model. Note that, for CICSPlex SM,
the dynamic routing program EYU9XLOP performs also the distributed routing function.
The advantage of the distributed model is that there is no single point of failure. The
disadvantages are:
v Compared with the hub model, there are a great many inter-region connections
to maintain.
v You may have problems with distributed data. For example, any data used to
make routing decisions must be available to all regions. With CICSPlex SM, this
problem is solved by the use of data spaces.
WLM functions
CICSPlex SM’s dynamic routing program supports:
v Workload separation: see page 63
v Workload balancing: see page 61
v Intertransaction affinity: see page 64
Workload balancing
Workload balancing is the routing of transactions or programs among a group of
target regions according to the availability and activity levels of those target regions.
Workload balancing can be used in addition to, or in place of, workload separation.
For example, CICSPlex SM can balance the workload among the SALESGRP
target regions by selecting, as each transaction is initiated, the target region that is
likely to deliver the best performance.
When selecting a target region, both the queue algorithm and the goal algorithm
take into account the way in which a target region is connected to its requesting
region. That is, a target region connected to its requesting region via MRO/XCF is
preferred to a VTAM®-connected target region, when all other considerations are
equal. For workload balancing of enterprise-bean related requests, only MRO
connections between routing regions and target regions are supported.
Workload balancing of enterprise beans can be achieved using the queue and goal
algorithms. The inbound IIOP work request is received by a routing region (listener)
and is matched to a bean name, an operation and a CorbaServer using a request
model definition. The routing region routes the transaction identified in the request
model to a target region. The transaction runs in the CorbaServer corresponding to
the installed request model instance.
This algorithm maximizes work throughput and standardizes response times across
the CICSplex. The queue algorithm is very robust: it can accommodate differences
in processor power; different maximum task values in the target regions;
asymmetric target region configurations; and an unpredictable workload.
MVS/ESA 5.1 (or later) is a prerequisite of the goal algorithm. Also, routing regions
in a CICSplex using the goal algorithm must be CICS TS regions.
The goal algorithm works best in environments where, if the CICSplex crosses
multiple MVS images:
v A similar number of target regions is on each of those MVS images.
# For example, if the default value of 65 is used, then all target regions on the CEC
# where this routing region resides must have a task load of 65% or higher before the
# routing region routes to target regions on other CECs, if all other health factors (for
# example, short-on-storage) for the routing regions are similar. As soon as target
# regions on other CECs achieve a task load of 65% or higher, the outing region
# resumes routing to the target regions on the local CEC again. When a local target
# region's task load drops below 65%, the routing region resumes routing to that
# target region regardless of the task load in target regions on remote CECs.
# Specifying this value lower than the default probably decreases the delay in routing
# to target regions on remote CECs. Take care not to set this value so low that the
# threshold is met by long-running tasks in the target regions.
# Specifying this value higher than the default most likely increases the delay in
# routing to target regions on remote CECs.
# Note that the effectiveness of this parameter is increased as the characteristics (for
# example, maxtask value or number of long running tasks) of the target regions
# become similar.
Workload separation
Workload separation routes work from a requesting or routing region among a set
of target regions, but the requirements of users, terminals, the transactions
themselves, and their BTS process type, can influence which target region set is
used. The routing of particular transactions or programs to a particular group of
target regions is based on any combination either of user ID, terminal ID, and
transaction group. For BTS processes, routing is based on a combination of
transaction group and process type. For enterprise bean-related transactions,
routing is based on transaction group. For Link3270 bridge requests, separating by
terminal luname is subject to some restrictions. See CICSPlex System Manager
Managing Workloads for more information.
Intertransaction affinity
An intertransaction affinity is a relationship between transactions, of a specified
duration, that requires them to be processed by the same target region. For
example, you might have a pseudoconversation made up of three separate
transactions, and each transaction passes data to the next transaction in the
sequence via a temporary storage queue (which may be shared in the distributed
model). You would then specify that all three transactions must be processed by the
same target region, and that this affinity lasts for the duration of the
pseudoconversation. (If you did not define this affinity to CICSPlex SM, each
transaction could be routed to a different target region and would therefore be
unable to access temporary-storage data left by the previous transaction.) The
target region itself is selected by CICSPlex SM from the specified target scope.
| Workload management and the IBM CICS Interdependency Analyzer for z/OS
understand affinities between BTS processes and activities. BTS itself does not
introduce affinities, and discourages programming techniques that do, but it does
handle existing code that may introduce affinities. You should define such affinities
to workload management, so that it can make sensible routing decisions. It is
particularly important to specify each affinity’s lifetime; failure to do so may restrict
unnecessarily workload management’s routing options.
| Workload management and the IBM CICS Interdependency Analyzer for z/OS do
not understand affinities between routable non-terminal-related EXEC CICS START
commands, or between DPLs not associated with a user id or a terminal. You
should take steps either to remove any affinities from your applications, or to ensure
that your applications honor any affinities.
Note that, if data is passed between transactions via the COMMAREA on the EXEC
CICS RETURN command, no such affinity exists: the COMMAREA is passed back
to the requesting region, and so can be passed to the target selected to process
the next transaction in the sequence. For more information about ways of avoiding
or minimizing transaction affinities, see the CICS/ESA publication Dynamic
Transaction Routing in a CICSplex.
WLMSCOPE
Routing Scope
WLMSPEC
Routing Algorithm
Affinity and Lifetime
Abend Compensation
Related RTA Event
Target Scope
WLMINSPC
Groups in Specification
WLMGROUP
WLMINGRP
Definitions in Group
Transaction Group
Status Transactions in Group
Terminal LUname
Affinity and Lifetime
User ID
Abend Compensation
Process Type
Related RTA Event
Target Scope
Having identified any affinities in a workload, you should make every attempt to
remove them. If you cannot remove them completely, gauge the duration of the
affinity and try to minimize it. You can define an intertransaction affinity to
CICSPlex SM as lasting:
v While the user’s session is active
v For the duration of the terminal session
v While the target region remains active
v While the workload is active
v For the duration of a pseudoconversation
v While the BTS activity is active
v While the BTS process is active
Be aware that CICSPlex SM must honor an active affinity: if an affinity is active but
the target region becomes unavailable, the transaction isn’t routed. In the case of a
BTS transaction, BTS will wait for the region to start.
Implementing WLM
What you have to do to use CICSPlex SM’s workload management functions varies
depending on whether you want to implement workload balancing only, or whether
you also want to implement some workload separation, and possibly define one or
more transaction affinities in the same CICSplex. You are recommended to follow
this process:
1. Identify candidates for dynamic routing
2. Implement workload balancing for each workload in the CICSplex where
dynamic routing is possible.
3. Add any necessary workload separation requirements.
4. Define any intertransaction affinities.
Workload balancing is effective within the workload when a routing region belonging
to the workload is next started.
Note: It’s important that you define routing regions and target regions in groups, to
ensure that changes to the composition of those groups are automatically
reflected in CICSPlex SM’s workload management.
A recommended approach
When you decide to implement WLM functions in your enterprise, you should begin
by defining the appropriate objects. You should:
v Use the WLMDEF view to create a workload definition.
You can use the MAP action command to display a visual map of the WLM
definitions in your data repository. If you want to see a list of the objects and
associations you have already defined, use the WLMINGRP, WLMINSPC, and
WLMSCOPE views.
If you have problems defining any particular object or association, refer to the
CICSPlex SM Starter Set examples. The Starter Set includes many WLM examples
that you can use as models for your own objects. You can copy the supplied
samples directly into your data repository via the batched repository-update facility.
For more information about using the Starter Set as a model see the CICS
Transaction Server for z/OS Installation Guide. For information about the batched
repository-update facility, see CICSPlex System Manager Administration.
Where next?
Once you have identified your WLM requirements, you should define them to
CICSPlex SM as described in CICSPlex System Manager Managing Workloads.
For example, if you tell CICSPlex SM that all files used by CICS systems
AORPAY1 must be in ENABLED status, CICSPlex SM sends out warning
messages when the status of any of those files is other than ENABLED. However,
CICSPlex SM isn’t restricted to telling you about a problem only after it has
occurred. CICSPlex SM’s real strength lies in its ability to detect potential problems
with CICS resources. Indeed, CICSPlex SM is unrivaled in its ability to detect and
warn of the slightest change in the status of a CICS resource. For example,
CICSPlex SM can tell you that resource contention is growing, or that the dynamic
storage area (DSA) free space is falling, or that the number of users of a particular
transaction is high: these can all be indicators of future problems that you could
avoid by taking preventive action.
The notification can take the form of a console message, or of a generic alert to
NetView, or both. The CICSPlex SM application programming interface (API)
makes it possible for automation products actively to gather status information from
CICSPlex SM, rather than waiting for CICSPlex SM to send out MVS messages
and SNA generic alerts.
Why automate?
The CICSPlex SM operator can monitor RTA messages and take appropriate
action. For example, the CICSPlex SM operator can change the status of a file
from DISABLED to ENABLED using the CICSPlex SM FILED view. However, this
approach is becoming less satisfactory as CICS availability and performance
requirements grow. Most enterprises are using more and more CICS systems, often
via cloning, to process a growing workload. This factor alone has made the
CICSPlex SM operator’s task more difficult and error prone. At the same time,
however, service-level agreements are setting performance and availability targets
that are more demanding than ever before. Consequently, the ideal environment is
one in which the CICSPlex SM operator is presented with problems only in the
most exceptional circumstances. Such an environment is now possible, because
most actual and potential problems detected by CICSPlex SM can be corrected or
prevented by intelligent use of automation products, such as:
v NetView
v Automated Operations Control/MVS (AOC/MVS)
© Copyright IBM Corp. 1994, 2010 73
v Automated Operations Control/MVS CICS Automation (AOC/MVS CICS
Automation)
v CICSPlex SM itself. (Not only does CICSPlex SM provide detailed and timely
information about CICS resource status to other automation products, it is also
capable of correcting many of the problems it detects.)
External messages
External messages are MVS/ESA operator messages that are written to the MVS
console by default. External messages can also be viewed from the CICSPlex SM
EVENT view while a condition is active. When a condition has been resolved, it
disappears from the EVENT view. (However, external messages written to the
EVENT view are also written to the CICSPlex SM EYULOG to provide a record of
the contents of the EVENT view.) In the EVENT view, external messages are sorted
in order of urgency, with the message about the most urgent condition at the top of
the list. The urgency of a condition is user-defined.
External messages are issued in pairs: one message is issued at the start of a
condition, and another is issued when the condition ends. The messages, which all
have the prefix “EYUPN”, are described in CICSPlex System Manager Messages
and Codes External messages can be intercepted by automation products, such as
NetView, AOC/MVS, and AOC/MVS CICS Automation.
RTA functions
The RTA functions are:
v System Availability Monitoring (SAM): see “System Availability Monitoring (SAM)”
on page 75
v RTA resource monitoring, which includes:
– MAS resource monitoring (MRM): see “MAS resource monitoring (MRM)” on
page 76
– Analysis point monitoring (APM): see “Analysis point monitoring (APM)” on
page 77
Advantages of RTA
The advantages of using RTA are:
v Actions can be taken within CICS, removing the requirement to use external
automation products.
v Potential problems are detected and corrected before they become critical.
v Performance to the end-user is improved by the early detection of bottlenecks
that could affect response time.
v The volume of information presented to operators is reduced by filtering out
unnecessary messages.
For example, if CICS system AORPAY2 must be active from 0900 hours to 1200
hours and CICSPlex SM finds it inactive at 0930 hours, CICSPlex SM issues
external messages by default. A typical response, from the CICSPlex SM operator
or from an automation product, would be to try to restart the CICS system, or
perhaps to start another.
You can customize the default SAM notifications for any CICS system by creating a
CICSPlex SM action definition. Using an action definition, you can change:
v The type of notification that is issued for a SAM condition. For example, you
could specify that, for a particular CICS system, the SAMSOS condition will
cause an SNA generic alert to be sent to NetView (so that NetView can take
corrective action), and that no external message is to appear in the EVENT view.
v The severity value and the priority value of a SAM condition. For example, you
could specify that the SAMTDM condition must have a severity of LW (Low
Warning) in a particular CICS system, and that the SAMSOS condition must have
a priority of 128.
You can also use the action definition to request that CICS TS systems affected by
a condition are to be cancelled and restarted using the MVS automatic restart
manager (ARM). This might be a suitable automation action for the SAMOPS
condition, for example.
However, most of the SAM notifications relate to actual problems that are already
having an effect on users, so your scope for preventive action is small.
Consequently, the best automation option for most SAM conditions is usually to
request a dump, which can provide more information about a problem and its
causes.
As its name suggests, MAS resource monitoring operates at the CICS system level.
That is, if a resource’s status changes in two regions, two sets of external
notifications are issued, one for each region.
When changes in the absolute status of a resource occur (for example, when a file
that should be ENABLED becomes DISABLED, or when a connection that should
be INSERVICE goes OUTSERVICE), remedial action is usually required. CICSPlex
SM itself can correct the status of such resources automatically, as can any of the
automation products, such as NetView or AOC/MVS CICS Automation.
Managing RTA
RTA is set up and managed using the objects described in Table 5.
Table 5. Real-time analysis objects
Object name Description Function
APACTV Analysis definitions associated with This object shown information about analysis definitions
an analysis point specification associated with an analysis point specification.
CICSSYS
CICSGRP
RTA Active
RTASCOPE
RTASPEC
SAM Action
SAM Severity
RTAINSPC
Groups in Specification
ACTNDEF
Event
Console Message
NetView Alert
ARM Restart RTAGROUP
RTAINGRP
Definitions in Group
PERIODEF
Active Period
RTADEF EVALDEF
Note: For APM only, you may also specify one or more secondary CMASs.
Implementing RTA
Real-time analysis should be implemented gradually and in response to an
established need for the resultant information, especially if you are not planning to
automate your responses to exception conditions. You are recommended to begin
by implementing SAM, and to implement MRM after having analyzed records, such
as problem logs, that give you some indication of where MRM would be useful.
These two actions are sufficient to activate SAM and to cause the default external
notifications to be generated when a noteworthy condition occurs. As you can see,
you can implement SAM as soon as CICSPlex SM is installed in your enterprise,
and with very little additional effort. You are recommended to continue using SAM
with the default notifications, at least until you have had an opportunity to see the
default notifications of all SAM conditions. If you decide to customize any of the
notifications, start by creating an action definition for one of the SAM conditions and
apply it to a single CICS system to test the effects. When you are happy with the
change, apply it to other SAM conditions and other CICS systems, as appropriate.
The overheads associated with SAM are not high, so you can activate it throughout
the enterprise CICSplexes without affecting the general performance of
CICSPlex SM or the CICS systems it manages.
The interval between resource evaluations (as specified in the evaluation definition)
and the intervals at which data is to be analyzed (as specified in the analysis
definition) should be intelligently matched (and ideally the same). In particular, do
not analyze the data more frequently than it is evaluated. Also, resource evaluations
should be performed as infrequently as possible while still providing useful data: the
more frequently data is collected, the greater the cost of its collection.
Begin by installing the analysis definition manually in the active CICS system. When
you are satisfied that MAS resource monitoring is working as you expected,
consider whether automatic installation of the analysis definition would be useful. If
it would, add the analysis definition to an analysis group, and associate the analysis
group with an analysis specification. At this point, decide whether to monitor
additional resources in the same CICS system. To extend MAS resource monitoring
to other resources, an additional evaluation definition would be required, an
additional analysis definition might be required, but the original action definition
could probably be reused.
A recommended approach
When you decide to implement RTA functions in your enterprise, you should begin
by defining the appropriate objects. You should:
v Use the RTADEF view to create an analysis definition.
v Use the EVALDEF and ACTNDEF views to create the required subcomponents
of an analysis definition.
v Use the RTAGROUP view to create an analysis group.
v Use the RTASPEC view to create an analysis specification.
Then, to create associations between these objects:
v Use the RTADEF view to add the analysis definition to the analysis group.
v Use the RTAGROUP view to add the analysis group to the analysis specification.
v Use the RTASPEC view to associate the analysis specification with a CICS
system or CICS system group.
You can use the MAP action command to display a visual map of the real-time
analysis definitions in your data repository. If you want to see a list of the objects
and associations you have already defined, use the RTAINGRP, RTAINSPC, and
RTASCOPE views.
If you have problems defining any particular object or association, refer to the
CICSPlex SM Starter Set examples. The Starter Set includes many real-time
analysis examples that you can use as models for your own objects. You can copy
the supplied samples directly into your data repository via the batched
repository-update facility. For more information about using the Starter Set as a
model and about the batched repository-update facility, see the CICS Transaction
Server for z/OS Installation Guide and CICSPlex System Manager Administration,
respectively.
Where next?
Once you have identified your RTA requirements, you should define them to
CICSPlex SM as described in CICSPlex System Manager Managing Resource
Usage.
What is monitoring?
CICSPlex SM monitoring supports the collection of performance-related data, at
user-defined intervals, for named resource instances within a set of CICS systems.
At user-defined intervals, monitoring gathers status information and statistics for
selected resources, wherever they exist in the CICSplex, and makes them available
at any point of control. CICSPlex SM gathers information using standard CICS
interfaces (EXEC CICS INQUIRE, EXEC CICS COLLECT STATISTICS) and can be
used instead of the CICS Monitoring Facility (CMF). If CMF is running,
CICSPlex SM exploits the information provided. Monitoring need not be specified
for resources being overseen by real time analysis (see “Implementing RTA” on
page 80), as these are monitored separately.
Monitoring functions
The CICSPlex SM monitoring functions are:
v Monitoring of resources: see “Monitoring of resources”
v Monitoring interval: see “The monitor interval” on page 84
v Sampling interval: see “The sample interval” on page 84
Advantages of monitoring
CICSPlex SM monitoring offers the following advantages:
v Summary and detailed information on resources is readily available, enabling you
to provide an improved help-desk service.
v System availability and performance are improved by allowing operators to be
more effective in investigating problems and seeking out bottlenecks.
Monitoring of resources
CICSPlex SM’s monitor functions can monitor the performance of these CICS
resources at specified intervals:
v CICS systems
v Connections
v DB2 threads
v FEPI connections
v Files and data tables
v CICS journals
v Programs
v Temporary storage
v Terminals
v Transactions
© Copyright IBM Corp. 1994, 2010 83
monitoring of resources
The data collected by this monitoring activity is either the type of data that is
collected by the CICS command EXEC CICS COLLECT STATISTICS, or is derived
from such data, that is, CICSPlex SM calculates rates, totals, averages, and
percentages from the raw statistics data. For local MASs only, CICSPlex SM can
also obtain some detailed monitoring data via the CICS monitoring facility (CMF).
CICSSYS
CICSGRP
MON active
MONSCOPE
MONSPEC
• MON active
• Data retention
• Sample intervals
• RODM CMAS
MONINSPC
Groups in specification
MONGROUP
MONINGRP
• Definitions in group
PERIODEF
• Active period
MONDEF
• Resource type
• Resource names
• Include or exclude
• RODM status
defaults to zero minutes, and so as soon as monitoring stops for any CICS
system, that system’s monitor data is lost. If some unexpected event
occurs, the data could be invaluable, so setting the retention period greater
than zero is advisable. Note, however, that if a CICS system stops, monitor
data is lost when you restart the system regardless of any retention period
you may have specified. In this case, it’s important to look at the monitor
data before you restart the CICS system.
You can set the retention period to any value between 1 and 1440 minutes
(24 hours). You can specify a retention period on the monitor specification,
in which case that period applies to every CICS system associated with the
monitor specification. Alternatively, you can specify a retention period in the
CICSPlex SM CICS system definition.
Decide when to activate monitor definitions
When you add a monitor definition to a monitor group, you can specify the
part of the day during which it is to be active. You should make considered
use of this feature. For example, you probably will not want monitoring
active overnight, or during any period when general use is low and no one
is around to view the data, so ensure that monitor definitions are not in
effect at those times. Also, if you’re monitoring resources for which there is
contention, think about when this contention occurs. Is it constant, or does it
occur only between 0900 and 1100 hours? If the latter, limit your monitoring
to that period.
You can refine your approach to monitoring gradually, by creating additional monitor
specifications for the CICSplex, for example, and by using timed monitor definitions,
which are installed automatically for a specified duration. Remember that:
v Although you can make changes to a monitor specification while it is in use, the
effect of those changes is not felt until the CICS system is next started. Note,
however, that you can use the MAS view to change monitoring values for an
active CICS system.
v A CICS system can be associated with only one monitor specification at a time.
v You can associate a monitor specification with a CICS system that is already
running, but the effect of that new association is not felt until the CICS system is
next started.
Aim to ensure that any monitor specifications and monitor definitions you provide
support the normal situation, and that CICS system overrides are used for the
exceptions.
For more information about the resource status facility see the description of the
RESSTATUS parameter in the CICS Transaction Server for z/OS Installation Guide
CMF data is usually written to an SMF data set. However, if you prefer, you can
collect CMF data for use by CICSPlex SM without having it written to an SMF data
set.
More information about the CICS SIT parameters and about suppressing the output
of CMF data to an SMF data set is provided in the CICS Transaction Server for
z/OS Installation Guide. For more information about CMF, see the CICS/ESA
Performance Guide.
A recommended approach
When you decide to implement MON functions in your enterprise, you should begin
by defining the appropriate objects. You should:
v Use the MONDEF view to create an analysis definition.
You can use the MAP action command to display a visual map of the MON
definitions in your data repository. If you want to see a list of the objects and
associations you have already defined, use the MONINGRP, MONINSPC, and
MONSCOPE views.
If you have problems defining any particular object or association, refer to the
CICSPlex SM Starter Set examples. The Starter Set includes many resource
monitoring examples that you can use as models for your own objects. You can
copy the supplied samples directly into your data repository via the batched
repository-update facility. For more information about using the Starter Set as a
model and about the batched repository-update facility, see CICS Transaction
Server for z/OS Installation Guideand CICSPlex SM Administration, respectively.
Where next?
Once you have identified your monitoring requirements, you should define them to
CICSPlex SM as described in CICSPlex System Manager Managing Resource
Usage.
Release levels are minimum levels only: later levels (if any) can be used.
v CICS Transaction Server for OS/390 Version 1 Release 3or later
| v The IBM CICS Interdependency Analyzer for z/OS can be used if you need
| assistance in identifying inter-transaction affinities
v ISPF Version 3 Release 3 is required to support the CICSPlex SM end-user
interface
v MVS/DFP Version 3 Release 1 is required
v MVS/ESA:
– MVS/ESA Version 5 Release 1 is required if the goal algorithm is to be used
for WLM
– MVS/ESA Version 5 Release 2 is required if the MVS/ESA automatic restart
manager (ARM) function is required
v NetView:
– CICSPlex SM can send generic alerts to NetView Version 2 Release 3, or
later
v RACF Version 2 Release 1 or equivalent external security manager is required
v TSO/E Version 2 Release 1 is required for support of ISPF
Prerequisite hardware
The basic requirement is for a processor that supports the prerequisite operating
system and has sufficient processor storage to meet the requirements of z/OS V1.4,
CICS TS for z/OS, Version 3.1, the application programs, the access methods, and
all other software being run. This includes the IBM eServer™ zSeries 990.
Storage requirements
See CICS Transaction Server for z/OS Program Directory for the minimum storage
requirements for CICSPlex SM on an MVS image on which a CMAS and one or
more MASs are defined.
The entity types, for each instance of which a name is required, are:
v CICSPlex SM components and CICSplex entities
– CAS, CMAS, CICSplex, MAS (CICS system), CICS system group, and
time-period definition.
v BAS entities
– Resource assignments, resource descriptions, resource definitions, and
resource groups.
Application resource definitions are:
- DB2 transactions, files, map sets, partition sets, programs, transient data
queues, transactions, CICS BTS process types, document templates, FEPI
nodes, FEPI pools, FEPI property sets, FEPI targets, file and key file
segment definitions, IIOP request models, map sets, sysplex enqueue
models, TCP/IP services, temporary storage models
Region resource definitions are:
- DB2 entries, enterprise beans, journals, journal models (CICS Transaction
Server for OS/390, Version 1 Release 3 or later), local shared resource
(LSR) pools, profiles, temporary storage queue models, transaction classes,
terminals, typeterms
System–system connectivity resource definitions are:
- Connections, DB2 connections, partners, sessions
Ensure that any convention you select can accommodate all of these entity types. A
naming convention, for use generally in CICS environments, is described in the
CICS/ESA publication Dynamic Transaction Routing in a CICSplex. For an example,
see “The Starter Set naming convention” on page 125.
Generic names
CICSPlex SM supports the use of generic names in many of its commands. For
example, you can specify that all transactions whose names begin with the letters
“DNW” are to be monitored once every 300 seconds. You do not have to name
each transaction individually.
Security planning
CICSPlex SM uses a SAF-compliant external security manager, such as RACF:
v To prevent unauthorized access to CICSPlex SM functions and CICS resources
v To control the simulation of CICS command checking and CICS resource
checking
In both cases, security checking is handled by the CMASs managing the CICS
systems that are the target of any request to access a resource. For example, if a
CICSplex is managed by two CMASs, and a request is made to access a resource
in all CICS systems belonging to that CICSplex, the security check is performed in
both CMASs.
To activate security checking, you must modify the JCL used to start the CMAS or
its managed CICS systems. If security checking is switched off for the CICS
system, no checking occurs, regardless of the CMAS setting. However, if security
checking is switched off for the CMAS but switched on for the CICS system, the
CICS system is not able to connect to the CMAS.
You should begin by deciding how much security checking you need. In particular,
identify those users who need access to CICSPlex SM, and ensure that an
individual user has the same user ID across all systems on which a CMAS is
installed. The user ID against which the security check is performed is the
signed-on TSO user ID. Consider also the type of security checking you want to
implement.
An external security manager is also used to protect CICSPlex SM’s own libraries,
procedures and Web User Interface resources. Full details of how to protect
CICSPlex SM’s libraries and procedures are provided in CICS RACF Security
Guide. In order to protect Web User Interface views, menus, help information and
the View Editor you need to create an approptiate profile in the FACILITY class.
See the CICSPlex System Manager Web User Interface Guide for more
information.
You should also take care if you are using the EXEC CICS CREATE command to
build new resources. Any definition created with the CICSplex as the context is
automatically distributed to all CMASs in the CICSplex. Therefore, giving a user
authority to create BAS objects is equivalent to giving authority to install resources
on any CICS system in the CICSplex. When the CICS system starts, there is no
check on who installed the resource in the system.
Time zones are specified using single-character codes in the range B through Z.
For example, code “S” represents Mountain Standard Time, code “T” represents
Central Standard Time, and code “C” represents Eastern Europe time. A complete
list of the codes can be found in CICSPlex SM Administration. CICSPlex SM allows
offsets (known as “time zone adjustments”) in the range 0 through 59 minutes to be
specified to accommodate regions that are not running in the standard time zones.
Also, daylight saving time can be specified.
| For example, a typical command file for input to the batched repository-update
| facility would contain a Context command, followed by multiple Create commands.
| An example command file is supplied with the CICSPlex SM Starter Set. Detailed
| information about the batched repository-update facility is supplied in CICSPlex SM
| Administration.
| Here are some examples that illustrate how CICSPlex SM handles variations in
| support for resources and functions throughout a CICSplex that includes CICS
| TS.systems.
| v Example 1: monitoring of transient data queues
| You create a monitor definition for transient data queues, add it to a monitor
| group, and associate the monitor group with a monitor specification. The scope
| of the monitor specification is every CICS system in the CICSplex. As each CICS
| system starts, CICSPlex SM determines whether it can install the monitor
| definition in that system. If not, CICSPlex SM issues a message to inform you
| that the monitor definition could not be installed. Both the monitor function and
| transient data queue resources are supported under CICS TS, so CICSPlex SM
| installs the monitor definition in those environments and monitoring begins.
| v Example 2: RTA of monitored programs
| You create an evaluation definition and name MPROGRAM (monitored programs)
| as the resource to be evaluated by CICSPlex SM’s RTA function. The evaluation
| definition is named in an analysis definition, which is linked, via an analysis
| group, to an analysis specification. The scope of the analysis specification is
| every CICS system in the CICSplex. The analysis definition is installed in the
| CICS TS systems. If an analysis definition names multiple evaluation definitions,
| CICSPlex SM installs those that are supported in the target environment.
Where next?
At this point, you are ready to install CICSPlex SM and run the JCL to install your
CASs, CMASs, and MASs. You now need to define the objects that will allow these
entities to communicate with CICSPlex SM and each other. CICSPlex SM
Administration gives you guidance on defining the CICSPlex SM configuration and
topology.
It is not intended that you should use the Starter Set as an initial configuration to
build on by replacing its data with your own. There is nothing to stop you doing this,
of course, but, as the Starter Set configuration is unlikely to resemble your own to
any significant extent, the benefits would be few. Also, if you update the Starter Set
definitions with your own data, you will effectively lose the Starter Set. New users of
CICSPlex SM would not have the opportunity to learn from the Starter Set, and
examples from the Starter Set, which are used throughout the CICSPlex SM
documentation, would be less valid. Therefore, you are recommended to retain the
Starter Set as a separate, sample configuration to which you can refer at any time,
but to copy anything (such as JCL or CICSPlex SM definitions) that you want to
adapt for local use.
CAS CAS
EYUCAS1A EYUCAS1B
CMAS CMAS
EYUCMS1A EYUCMS1B
Figure 22. The Starter Set configuration. The Starter Set provides a small but comprehensive set of CICSPlex SM
components on which you can model your own CICSPlex SM environment.
This JCL has been provided chiefly so that you don’t have to create it yourself
before you can use the Starter Set. However, you might also find it useful to refer to
when you are creating JCL for your own CICSPlex SM environment (for which you
are recommended to use the skeleton JCL supplied with CICSPlex SM).
CICSTS31.CPSM.SEYUJCL also includes sample JCL for deleting the Starter Set
components from an MVS image.
There are two IVPs, which are referred to as IVP1 and IVP2. IVP1 is run on the first
(or only) MVS image on which CICSPlex SM is installed and uses the system A
components of the Starter Set. IVP2 is run on the second and subsequent MVS
images on which CICSPlex SM is installed. It uses both the system A and system
B components of the Starter Set, and tests communication between CASs and
CMASs. The IVPs are described in detail in the CICS Transaction Server for z/OS
Installation Guide.
You add some of these definitions to the CMAS data repository interactively when
you follow the instructions for configuring the Starter Set in the CICS Transaction
Server for z/OS Installation Guide. However, you add most of them to the CMAS
data repository using the CICSPlex SM batched repository-update facility. A brief
description of all of the Starter Set definitions is supplied in Table 7 on page 105
through Table 54 on page 124. For a more detailed understanding of these
definitions, you are recommended to study the definitions as supplied in
CICSTS31.CPSM.SEYUDEF.
Figure 23. A Starter Set topology view. CICSPlex SM view SYSGRPC lists the CICS
systems in a specified scope. This example lists the CICS system groups in EYUPLX01, and
identifies the CICS systems and CICS system groups that belong to those four groups.
BAS definitions
The Starter Set BAS definitions include resource definitions, communication links,
resource groups, resource assignments, and resource descriptions.
Table 16. Resource group and resource definitions for CPUONLY application
Name Belongs to Description
EYUBAG02 N/A Identifies the resource group for the CPU-only
application.
ET01 EYUBAG02 Transaction definition (for SSET) that calls the
EYUPR001 program.
ET02 EYUBAG02 Transaction definition (for SSET) that calls the
EYUPR002 program.
EYUPR001 EYUBAG02 Program definition (for SSET) called by transaction
ET01.
EYUPR002 EYUBAG02 Program definition (for SSET) called by transaction
ET02.
Table 19. Resource group and resource definitions for autoinstall programs
Name Belongs to Description
EYUBAG05 N/A Identifies the resource group for the autoinstall
programs.
Figure 24. A Starter Set BAS view. CICSPlex SM view RASGNDEF shows the
characteristics of the sample resource assignments used to assign selected resource
definitions to CICS systems. For each resource assignment, the resource definitions to be
assigned are of a single resource type and are associated with a resource group. A resource
assignment must be associated with at least one resource description (RESDESC) before
any assignment can begin, as shown in Figure 25 (RASINDSC view).
Figure 25. A Starter Set BAS view. CICSPlex SM view RASINDSC lists the sample resource
assignments and their associated resource descriptions.
Figure 26. A Starter Set workload management view. This example shows the WLMSPEC
view, which is displaying the workload specifications for CICSplex EYUPLX01.
Figure 27. A Starter Set monitoring view. The MONDEF view displays all monitor definitions
for CICSplex EYUPLX01.
Figure 28. A Starter Set RTA view. This example shows the RTAGROUP view, which is
displaying the analysis groups in EYUPLX01.
The status definitions supplied with the Starter Set are triggered by CICS programs
known as status probes. These programs are shipped in source format in data set
CICSTS31.CPSM.SEYUDEF. You are recommended to read the commented source
files before using any of these programs. The status probe programs and the
definitions that use them are:
It is most important that you define an appropriate naming convention for your own
CICSPlex SM environment. How to approach this is described in Chapter 9,
“Planning CICSPlex SM installation and setup,” on page 91.
PDF-only books
The following books are available in the CICS Information Center as Adobe
Portable Document Format (PDF) files:
Licensed publications
The following licensed publications are not included in the unlicensed version of the
Information Center:
CICS Diagnosis Reference, GC34-6899
CICS Data Areas, GC34-6902
CICS Supplementary Data Areas, GC34-6905
CICS Debugging Tools Interfaces Reference, GC34-6908
Subsequent updates will probably be available in softcopy before they are available
in hardcopy. This means that at any time from the availability of a release, softcopy
versions should be regarded as the most up-to-date.
For CICS Transaction Server books, these softcopy updates appear regularly on the
Transaction Processing and Data Collection Kit CD-ROM, SK2T-0730-xx. Each
reissue of the collection kit is indicated by an updated order number suffix (the -xx
part). For example, collection kit SK2T-0730-06 is more up-to-date than
SK2T-0730-05. The collection kit is also clearly dated on the cover.
Bibliography 129
130 CICS TS for z/OS: CICSPlex SM Concepts and Planning
Accessibility
Accessibility features help a user who has a physical disability, such as restricted
mobility or limited vision, to use software products successfully.
You can perform most tasks required to set up, run, and maintain your
CICSPlex SM system in one of these ways:
v using a 3270 emulator connected to CICSPlex SM
v using a 3270 emulator logged on to CICS
v using a 3270 emulator logged on to TSO
v using a 3270 emulator as an MVS system console
v using the CICSPlex SM web user interface.
IBM Personal Communications (Version 5.0.1 for Windows 95, Windows 98,
Windows NT and Windows 2000; version 4.3 for OS/2) provides 3270 emulation
with accessibility features for people with disabilities. You can use this product to
provide the accessibility features you need in your CICSPlex SM system.
Index 135
software requirements 91 view (continued)
software, prerequisite 91 BATCHREP 97
staged implementation 34 CASACT 35
Starter Set CASDEF 35
BAS samples 107 CASDEFD 35
CAS definitions 104 CICSGRP 39
configuration samples 105 CICSPLEX 36
definitions, supplied 104 CICSSYS 39
installation requirements 102 CMAS 36
JCL, supplied 103 CMASD 36
monitoring samples 113 CMASPLEX 36
naming convention 125 CMTCMDEF 36
purpose of 101 CMTCMLND 36
RTA samples 116 CMTCMLNK 36
structure of 101 CMTPMLND 36
topology samples 105, 106 CMTPMLNK 36
workload management samples 111 CONNDEF 39, 46
starting CICSPlex SM 96 CPLEXDEF 36
STATDEF view 78 CPLXCMAS 36
storage requirements 92 DB2CDEF 46
SYSGRPC view 39 DB2EDEF 46
SYSLINK view 40, 46 DB2TDEF 46
system availability monitoring (SAM) DTRINGRP 65
automation and 76 EVALDEF 78
customizing external notifications 75 FENODDEF 46
description of 75 FEPOODEF 46
implementing 80 FEPRODEF 46
system group 21, 22 FETRGDEF 46
FILEDEF 46
FSEGDEF 46
T HTMLDEF 46
TCPDEF view 46 JRNLDEF 46
TDQDEF view 46 JRNMDEF 46
TERMDEF view 46 LSRDEF 46
time zone 6, 94 MAPDEF 46
Tivoli Business Systems Manager MAS 40
support for 15 MONDEF 84
topology 2 MONGROUP 84
administration 38 MONINGRP 84
topology definitions MONINSPC 85
Starter Set 105, 106 MONSCOPE 85
TRANDEF view 46 MONSPEC 85
TRANGRP view 65 PARTDEF 46
TRNCLDEF view 46 PERIODEF 39, 80, 94
TSMDEF view 46 PLEX 35
TYPTMDEF view 46 PLEXOVER 35
PROCDEF 46
PROFDEF 46
U PROGDEF 46
unauthorized access to CICSPlex SM, preventing 94 PRTNDEF 46
updating the configuration 34 RASGNDEF 45
RASINDSC 45
RASPROC 45
V RDSCPROC 46
RESDESC 46
validating resources 49
RESGROUP 46
view
RESINDSC 46
ACTDEF 78
RESINGRP 46
ADMCONFG 36
RQMDEF 46
ADMTOPOL 38
RTADEF 78
APCMAS 78
RTAGROUP 78
APSPEC 78
W
Web User Interface 7, 11
WLMDEF view 66
WLMGROUP view 66
WLMINGRP view 66
WLMINSPC view 66
WLMSCOPE view 66
WLMSPEC view 66
workload definition 66
Starter Set 111
workload functions of CICSPlex SM 57
workload group 66
workload management (WLM)
definition of 4
intertransaction affinity 61, 64
introduction to 57
Starter Set samples 111
workload balancing 61, 69
abend compensation 62
goal algorithm 62
queue algorithm 62
workload separation 63, 70
workload specification 66
Index 137
138 CICS TS for z/OS: CICSPlex SM Concepts and Planning
Notices
This information was developed for products and services offered in the U.S.A. IBM
may not offer the products, services, or features discussed in this document in other
countries. Consult your local IBM representative for information on the products and
services currently available in your area. Any reference to an IBM product, program,
or service is not intended to state or imply that only that IBM product, program, or
service may be used. Any functionally equivalent product, program, or service that
does not infringe any IBM intellectual property right may be used instead. However,
it is the user's responsibility to evaluate and verify the operation of any non-IBM
product, program, or service.
IBM may have patents or pending patent applications covering subject matter
described in this document. The furnishing of this document does not give you any
license to these patents. You can send license inquiries, in writing, to:
For license inquiries regarding double-byte (DBCS) information, contact the IBM
Intellectual Property Department in your country or send inquiries, in writing, to:
The following paragraph does not apply in the United Kingdom or any other
country where such provisions are inconsistent with local law:
INTERNATIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS
PUBLICATION “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS
OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
OF NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A
PARTICULAR PURPOSE. Some states do not allow disclaimer of express or
implied warranties in certain transactions, therefore this statement may not apply to
you.
Licensees of this program who wish to have information about it for the purpose of
enabling: (i) the exchange of information between independently created programs
and other programs (including this one) and (ii) the mutual use of the information
which has been exchanged, should contact IBM United Kingdom Laboratories,
MP151, Hursley Park, Winchester, Hampshire, England, SO21 2JN. Such
information may be available, subject to appropriate terms and conditions, including
in some cases, payment of a fee.
Trademarks
IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of
International Business Machines Corp., registered in many jurisdictions worldwide. A
current list of IBM trademarks is available on the Web at Copyright and trademark
information at www.ibm.com/legal/copytrade.shtml.
Other product and service names might be trademarks of IBM or other companies.
Feel free to comment on what you regard as specific errors or omissions, and on
the accuracy, organization, subject matter, or completeness of this book.
Please limit your comments to the information in this book and the way in which the
information is presented.
To ask questions, make comments about the functions of IBM products or systems,
or to request additional publications, contact your IBM representative or your IBM
authorized remarketer.
When you send comments to IBM, you grant IBM a nonexclusive right to use or
distribute your comments in any way it believes appropriate, without incurring any
obligation to you.
You can send your comments to IBM in any of the following ways:
v By mail, to this address:
IBM United Kingdom Limited
User Technologies Department (MP095)
Hursley Park
Winchester
Hampshire
SO21 2JN
United Kingdom
v By fax:
– From outside the U.K., after your international access code use
44–1962–816151
– From within the U.K., use 01962–816151
v Electronically, use the appropriate network ID:
– IBMLink: HURSLEY(IDRCF)
– Internet: idrcf@hursley.ibm.com
SC34-6459-03
Spine information:
Version 3
CICS TS for z/OS CICSPlex SM Concepts and Planning Release 1